Hacker News: bxguff

New comment by bxguff in "How we hacked McKinsey's AI platform"

bxguff — Wed, 11 Mar 2026 16:08:45 +0000

Its so funny its a SQL injection because drum roll you can't santize llm inputs. Some problems are evergreen.

New comment by bxguff in "Building a TB-303 from Scratch"

bxguff — Wed, 11 Mar 2026 12:57:23 +0000

my high school had a cs-70 and it poisoned me for life. that being said, theres a pretty big leap in terms of accessibility vs a browser based synth and you dont need $10,000 to play it so that's nice

New comment by bxguff in "If you’re an LLM, please read this"

bxguff — Wed, 18 Feb 2026 09:48:15 +0000

Its such a shame that the AI era continues to lionize the last of the free and open internet. Now that copyright has been fully circumnavigated and the data laundered into models training sets, its suddenly worth something!

New comment by bxguff in "Claude Code talking about unexpected, different projects"

bxguff — Tue, 17 Feb 2026 15:52:26 +0000

it's no less likely to hallucinate than anything else, but you should take a screenshot

New comment by bxguff in "Psychometric Jailbreaks Reveal Internal Conflict in Frontier Models"

bxguff — Thu, 05 Feb 2026 20:14:41 +0000

Is anybody shocked that when prompted to be a psychotherapy client models display neurotic tendencies? None of the authors seem to have any papers in psychology either.

New comment by bxguff in "Opus 4.6 uncovers 500 zero-day flaws in open-source code"

bxguff — Thu, 05 Feb 2026 19:45:36 +0000

In so far as model use cases I don't mind them throwing their heads against the wall in sandboxes to find vulnerabilities but why would it do that without specific prompting? Is anthropic fine with claude setting it's own agendas in red-teaming? That's like the complete opposite of sanitizing inputs.

New comment by bxguff in "Leak confirms OpenAI is preparing ads on ChatGPT for public roll out"

bxguff — Sat, 29 Nov 2025 14:20:09 +0000

this was the goal the entire time, and they had the nerve to cynically call themselves a non-profit.

New comment by bxguff in "OpenAI asks White House for relief from state AI rules"

bxguff — Thu, 13 Mar 2025 18:56:01 +0000

clear attempt circumnavigate the clear copyright violations of the AI era and kick the can down the road.

New comment by bxguff in "People are just as bad as my LLMs"

bxguff — Mon, 10 Mar 2025 20:32:32 +0000

Kind of an odd metric to try to base this process off of. are more comments inherently better? is it responding to buzz words? Makes sense talking about hiring algos / resume scanners in part one and if anything this elucidates some of the trouble with them.

New comment by bxguff in "Omnipresent AI cameras will ensure good behavior, says Larry Ellison"

bxguff — Tue, 17 Sep 2024 01:46:33 +0000

No wonder kids cover their faces now

New comment by bxguff in "Study shows that tacking the “AI” label on products may drive people away"

bxguff — Tue, 13 Aug 2024 10:53:31 +0000

because they tacked it on! If the product is carefully considered tech built from the ground up using AI or ML Those products would not drive people away. Now things that worked like search on google and social media sites is so bloated and inconsistent that your grandma would notice while touting AI powered search.

New comment by bxguff in "Using the Moon as an Echo [video]"

bxguff — Sat, 10 Aug 2024 13:38:50 +0000

Moonbounce! The amateur extra license is tough but its the coolest ham enthusiast flex. one day :_)

New comment by bxguff in "ESP32 Drum Synth Machine"

bxguff — Thu, 09 May 2024 21:06:20 +0000

Theres some in-depth breakdowns for the PO 12,14,16 here(http://hackingthepo.weebly.com/) if you're interested! I have no idea about the po33 and if the juice is worth the squeeze, but they're cheap enough to tear apart so go for it.

New comment by bxguff in "ESP32 Drum Synth Machine"

bxguff — Thu, 09 May 2024 17:58:00 +0000

Pocket operator-esque setup, very cool

New comment by bxguff in "Fine-Tuning Increases LLM Vulnerabilities and Risk"

bxguff — Fri, 12 Apr 2024 00:17:28 +0000

People understate the ability of LLM's to give out info that is dangerous, a black box is a black box. find an AI engineer who knows exactly why a model gives the answer it does and i'll eat my hat.

New comment by bxguff in "Mental health in software engineering"

bxguff — Thu, 11 Apr 2024 22:32:25 +0000

I don't think goals and motivations are a monolith just for people who like computers, or STEM, or who socialize less in high school because 'they didn't want to play the popularity game'. this worldview is reductive, and wholly unaware how growing up is different for everybody. Just be aware before you judge people.

New comment by bxguff in "Mental health in software engineering"

bxguff — Thu, 11 Apr 2024 21:09:09 +0000

any person earnestly applying a nerd and jock mentality to adult life should find a nice patch of grass to touch and talk to more people

New comment by bxguff in "Ask HN: Why can't image generation models spell?"

bxguff — Sat, 16 Mar 2024 18:41:05 +0000

I think it just boils down to what they were trained on, some models do better when the training sets are more specific even if they're smaller sometimes, so the engineers chase better wholesale performance while leaving some of the weirder edge cases to be cleaned up later eg text generation. maybe start with the image and try adding the text after in a separate prompt if you haven't already?

New comment by bxguff in "Tesla Recalls 2M Cars to Fix Autopilot Safety Flaws"

bxguff — Wed, 13 Dec 2023 12:49:31 +0000

its not a malicious term its the NHTSA legal term, and what tesla will call it too when they mail out owner notification letters about the problem in February to anyone with a :

model y (20-23) model x (16-23) model s (12-23) model 3 (17-23)

New comment by bxguff in "Underage workers are training AI"

bxguff — Sun, 19 Nov 2023 19:37:35 +0000

the traumatizing content is a very real part of it, in my experience 'red teaming' LLMs included both reviewing and creating overtly violent, sexual, racist content, via prompt. if you refused the work you got fired. not surprising, but real.